AI Con USA 2026 - AI Infrastructure
Monday, June 8
A Quality Engineering Introduction to AI and Machine Learning
Although there are several controversies and misunderstandings surrounding AI and machine learning, one thing is apparent — people have quality concerns about the safety, reliability, and trustworthiness of these types of systems. Not only are ML-based systems shrouded in mystery due to their largely black-box nature, they also tend to be unpredictable since they can adapt and learn new things at runtime. Validating ML systems is challenging and requires a cross-section of knowledge, skills, and experience from areas such as mathematics, data science, software engineering, cyber-security,...
Tuesday, June 9
Strategies for Testing Autonomous AI and Multi-Agent Architectures
NewTesting Artificial Intelligence (AI) agents presents a paradigm shift from traditional software quality assurance. Unlike deterministic, rule-based applications, AI agents exhibit emergent behaviors, learn from their environments, and make autonomous decisions, making conventional test case design and execution insufficient. This tutorial will provide a comprehensive understanding of the unique challenges and advanced strategies required to effectively test single and multi-agent AI systems. Participants will learn how testing agents differ significantly from testing traditional software....
AI Deep Dive: Exploring AWS Using Real-World Scenarios
Deepen your AI and machine learning expertise using AWS in an Immersive, hands-on workshop. You’ll use real-world AI challenges while leveraging AWS services like Amazon SageMaker, Bedrock, and Lambda to build and optimize AI-driven solutions. As the session unfolds, new constraints and data anomalies will emerge, mirroring the complexities of real-world AI/ML implementation. Gain insight into how AI solutions perform under evolving conditions, learning to adapt, optimize, and troubleshoot unexpected challenges. Learn the importance of collaboration, strategic thinking, problem-solving,...
Agentic AI: From Rules to Reasoning
NewAI agents have existed for decades, but generative AI has fundamentally changed what agents can do and how they are designed and built. Come and explore the evolution of AI agents across two major waves. First, learn the foundations of Agentic AI through agents built using rules, heuristics, and traditional machine learning, examining where these approaches excel and why they struggle with complexity, ambiguity, and scale. Then dive into the second wave of agents powered by generative AI and multimodal large language models. These modern agents can reason, plan, use tools, and interact...
Wednesday, June 10
Architecting the Digital Work System: The Shift to Multi-Agent Enterprise Workflows
The era of single-turn AI prompting is over, giving way to the sophisticated "digital assembly line." This session guides technical leaders, CTOs, and AI product managers through the critical leap from isolated AI tools to coordinated, multi-agent constellations capable of autonomously executing complex enterprise workflows. Marshall will dissect the four levels of agentic evolution, focusing on today's epicenter of innovation: cross-system orchestration. Because the risk of cumulative failure multiplies with every automated handoff, he will unpack the architectural strategies required to...
Containers That Think: Building AI-Powered Self-Healing Applications That Never Go Down
Enterprise containerized applications face a critical reliability crisis with complex failure modes including memory leaks, cascading failures, network partitions, and resource contention that traditional monitoring tools cannot predict or resolve fast enough. Organizations typically experience multiple production incidents monthly with multi-hour resolution times that consume significant engineering resources while causing customer-facing outages and revenue loss. Traditional approaches rely on reactive monitoring, manual troubleshooting across distributed container environments, and time...
Tracing the Mind of the Machine: Observability for AI Agents
AI agents have evolved beyond LLM chatbots; they possess the ability to plan, reason, and act autonomously. However, as their autonomy increases, understanding how they make decisions becomes more challenging. Traditional methods of observability—such as metrics, logs, and traces—capture outcomes but do not reveal the underlying reasoning. This session will explore how AI Agent Observability can shed light on the decision-making process by collecting and analyzing agent traces. We will discuss emerging standards like the Model Context Protocol (MCP), which provides structured and shareable...
Revolutionizing Healthcare with AI
Unlock Exponential Productivity: The AI Maturity Model for Product Engineering
Are you ready to transform your productivity from incremental gains to exponential growth? Whether you're an individual contributor or engineering leader, this session introduces the AI Maturity Model, a proven framework that guides technical teams through five distinct levels of AI adoption—from 33% productivity boosts to an extraordinary 1000% increase. Discover how to navigate each stage: Level 1 (Foundation, 33%) builds essential AI awareness; Level 2 (Literacy, 75%) develops practical AI skills; Level 3 (Fluency, 300%) masters AI-assisted workflows; Level 4 (Agents, 500%) implements...
Energy-Efficient AI: Building Sustainable Data Pipelines for the Future
Artificial Intelligence is driving innovation across industries, but its growing energy demands pose critical challenges around cost, scalability, and sustainability. In this talk, Bhanu will share practical strategies for designing energy-efficient AI systems, focusing on: dynamic batching & KV caching for reducing inference overhead, sparse neural networks & structured pruning for lightweight models, carbon-aware scheduling to align compute with renewable energy, federated learning & edge deployments to reduce data transfer energy, and a sustainability maturity model for...
Why Your Agentic PR Never Gets Approved—Going from Vibe Coding to Agentic Engineering
You can write 100% of your code with AI, and only be 15% more productive. Want to know why? Solving the problem of moving faster as an engineer is about way more than code. You design, you problem solve, you discuss and communicate, you test, you shepherd the PR through review, you deploy, and you observe. So why are you being told you should be 10x faster now that AI can write code? David will discuss practical solutions for safely accelerating your entire workflow with AI. He will look at how to collaborate with PM, how to use AI planning, how to ensure validations are solid and builds...
Thursday, June 11
Engineering AI Infrastructure for Efficient Inference at Scale
As AI models grow in complexity and scale, inference efficiency has emerged as a critical engineering challenge for enterprise deployment. Traditional infrastructure built for training workloads often fails to meet the latency, throughput, and cost demands of large-scale inference operations. In this session, Sandeep will be sharing practical insights from engineering AI infrastructure at Broadcom, focusing on the end-to-end optimization of compute, networking, and storage subsystems. The talk explores techniques such as dynamic workload placement, adaptive batching, model quantization,...
From Reactive to Proactive: Intuit's Journey in AI-Powered Incident Management
With thousands of Intuit services, managing incidents efficiently and effectively becomes essential. In this talk, Akshay will examine how they have moved beyond traditional incident management methods to embrace AI-driven solutions for speed, resilience, and continuous improvement. The session will outline how Intuit has scaled incident management across thousands of services, ensuring robust support across their ecosystem. See how the team utilized real-time insights and advanced automation to enhance AWS-related incident detection and resolution across Intuit's infrastructure, as well...
User-centricity for AI-assisted Test Engineers
Traditional software testing is fundamentally deterministic: the same inputs must always produce the same outputs. Yet many teams introduce AI into their testing without first defining the problem the AI is meant to solve, leading to brute-force experimentation and unreliable results. Google’s 2025 DORA report highlights that user-centricity is a prerequisite for AI success and that AI is most effective when it is pointed at a clear problem. AP shows you how that insight applies technically to testing. Before AI can be used as a testing tool, it must first be tested and understood in the...